Exponential Random Graph Modeling of Co-Offender Drug Crimes

• Main Authors: Tsai, Fu-Ching, Hsu, Ming-Chun, Kao, Da-Yu
• Format: Conference Proceeding
• Language: English
• Subjects: drug crime; Drugs; Education; ERGMs; Estimation; Law enforcement; Peer-to-peer computing; social network analysis; Social network services; Task analysis
• ISSN: 1738-9445
• DOI: 10.23919/ICACT48636.2020.9061552

Drug problem has contributed a rapid impact on today's society. It is not only a threat of human health but also causes a great impact on the social security issue. As drug abuse tends to organized crimes, we must consider the social network relationships among criminals to formulate better strategies against drugs. This research applied exponential random graph models (ERGMs) to analyze dynamic relations of drug crime. The strength of ERGMs is the ability to handle complicated dependency patterns which violate the basic assumption of traditional statistical methodologies. The homophily test and Monte Carlo Markov Chain (MCMC) estimation are used to explore the drug offenders' attributes and structural interactions, respectively. The experimental result shows that the homophily effect is significant on drug co-offenders relations regarding to occupation, education, nationality, drug type and recidivism. In addition, the positive 2-path coefficient indicates that drug offenders tend to share friends and form a cluster. The results of this paper reveal the advantages of structural implications in analyzing drug-related crime, as well as its ability to facilitate the cognition of crime prevention and intervention strategies.